"Monitoring Cavitation in a Cetrifugal Pump" by Julia Radcliffe and Gil Guinto
 

Monitoring Cavitation in a Cetrifugal Pump

Level of Education of Students Involved

Undergraduate

Faculty Sponsor

Shahin Nudehi

College

College of Engineering (COE)

Discipline(s)

Engineering

Presentation Type

Poster Presentation

Symposium Date

Spring 4-24-2025

Abstract

Our goal is to develop a real-time cavitation detection system for centrifugal pumps that prevents damage without requiring prohibitively low pressures. We use acoustics for detection by recording the noise produced by cavitation with an Integrated Circuit Piezoelectric (ICP) microphone (130D20, PCB Piezotronics) and analyzing the frequency domain of the data. Computational Fluid Dynamics (CFD) models are created at different inlet pressures to compare empirical results and understand pressure drops in the turbine. Data is acquired using a National Instruments Data Acquisition (NI DAQ) system and Laboratory Virtual Instrument Engineering Workbench (LabVIEW) software on a PC. We hypothesize that cavitation is signaled by a large magnitude of noise at a specific frequency.

The study uses a two-inch polyvinyl chloride (PVC) pipe that carries water from a centrifugal pump to a heat pump tank. Initially, the microphone is held by hand, but later it is fixed using a 3D-printed fixture bolted around the pipe. We collect multiple 5-second sound recordings at pressure differences ranging from 5 pounds per square inch (PSI) to 50 PSI, with the test pressure controlled via a butterfly valve and pressure gauge. The recorded data is analyzed in Matlab for deconvolution, comparing normal pressure conditions (5 PSI) to different pressure deviations.

The analysis shows that cavitation can be detected acoustically, with distinct peaks in the Fast Fourier Transforms (FFTs) of deconvolved samples at higher pressure deviations. We then develop a Matlab program that identifies cavitation by searching for repeated peaks in the deconvolved frequency domain plots. Cavitation is detected both qualitatively and quantitatively across multiple frequencies. The system successfully detects cavitation, preventing damage and serving as an educational tool.

Biographical Information about Author(s)

Gil Guinto: I am a transfer student, pursuing mechanical engineering in my Junior year, and research has been the focal point of my education career as a means to figure out what I want to do if I decide to pursue graduate school. As such, I have been taking up opportunities in research like GE190 to better equip myself with the knowledge and tools needed for future research. In this class I was presented with a project in the realm of fluid dynamics; cavitation. Which has been a topic of interest of mine after ME373 Fluid Mechanics. Within the project I am doing the CFD models that work in conjunction with ME476 CFD, as I am also doing research on cavitation within centrifugal pumps. (Graduating, Gil Guinto; 2026) (ORCID ID: https://orcid.org/0009-0004-9990-7565)

I’m Julia Radcliffe, a senior computer engineering student at Valparaiso University. I grew up on a farm in Northeast Indiana where my love for engineering all began. I’ve always been fascinated by how things work and plan to pursue a career in embedded software and application engineering. I’m especially passionate about machine learning, simulations, and modeling. As I wrap up my studies, I’m excited to take what I’ve learned and apply it in the real world.

This document is currently not available here.

Share

COinS